discounting disaster: land markets and climate change in ...€¦ · the sundarbans landscape is...

Discounting Disaster:

Land Markets and Climate Change in the Indian Sundarbans

.

Sumana Bandyopadhyay* Professor, Department of Geography, University of Calcutta

Sunando Bandyopadhyay Professor, Department of Geography, University of Calcutta

Susmita Dasgupta Lead Environmental Economist, Development Economics Research Group, World Bank

Chinmoyee Mallik

Lecturer, Birati College

David Wheeler Senior Fellow, World Resources Institute

October 2019

This research was conducted under the South Asia Water Initiative - Sundarbans Landscape

administered by the World Bank.

* Authors’ names in alphabetical order.

Corresponding Author: Sumana Bandyopadhyay. Professor, Department of Geography,

University of Calcutta, 35 Ballygunge Circular Road, Kolkata 700019, India.

Email: [email protected];

Tele phone: +91 9433015697; Fax:

Disclaimer of the World Bank: The findings, interpretations, and conclusions expressed in this

paper are entirely those of the authors. They do not necessarily represent the views of the

International Bank for Reconstruction and Development/World Bank and its affiliated

organizations, or those of the Executive Directors of the World Bank or the governments they

represent.

1

Abstract

In this paper, we use new household survey and environmental data to investigate the

relationships linking land prices, household incomes and climate-change-related factors in the

Indian Sundarbans. Although empirical work on land prices and climate change factors has

begun in developed countries, particularly the United States, we believe that this is the first such

exercise for the coastal region of a developing country. Our household data are drawn from a

georeferenced survey of land transactors that includes information on plot sizes and prices, as

well as the incomes and other characteristics of transactor households. We quantify

environmental conditions using two data sources: a georeferenced panel of salinity measures

drawn from 342 monitoring stations in the Indian Sundarbans region, and a spatial raster

database that incorporates information on all cyclonic storms that have struck the region since

1970.

Using these data, we treat land prices and household incomes as jointly-determined variables in

an econometric model that incorporates the effects of salinization, cyclonic storm intensity, and

inundation risk from proximity to the coastline. For both land prices and incomes, we find that

the estimated parameters for the environmental variables have the expected signs and generally-

high levels of significance. We use the regression results to predict land prices and household

incomes for cases where salinization, cyclonic storm intensity and inundation risk are at their

least and most favorable settings in the sample. We find very large impacts, with least and most

favorable cases typically yielding land prices and incomes below the sample 10th percentile and

near the 90th percentile, respectively.

We draw two broad conclusions from our results. First, current climate-change-related

conditions account for spatial differentials greater than an order of magnitude in Sundarbans land

prices and incomes. By implication, high-salinity, cyclone-prone areas near the coastline are

zones of poverty in the Sundarbans, while low-salinity areas far from the coastline with minor

cyclone histories are zones of relative affluence. Our second conclusion is driven by current

trends, which suggest that salinization, inundation risk, and cyclone strike frequency and

intensity will increase steadily as climate change proceeds. As this happens, our results suggest

that the most-affected areas will be abandoned by all but the poorest residents, who will be the

last to flee as the sea comes steadily inland.

2

1. Introduction

The Sundarbans Landscape is the UNESCO Heritage mangrove forest that extends across the

India-Bangladesh border at the mouth of the Ganges-Brahmaputra-Meghna river basin. The

incidence of poverty is strikingly high in the Sundarbans, and poor people in the region rely

mostly on natural resources for their livelihoods. Living conditions are harsh in this flat

landscape. Soil and water salinity are problems during annual peak periods; salinization is

spreading inland as sea level rise continues; and areas near the sea are increasingly prone to tidal

surges and cyclones. Poverty, topography, salinization and rising inundation risk have combined

to put the Sundarbans on the leading edge of climate-related problems in coastal regions of the

developing world.

As the sea rises and salinity spreads inland, coastal communities will confront inexorably-rising

pressure to adapt or relocate (Dasgupta et al. 2016). Their responses are bound to vary

considerably, because their decisions will be subject to uncertainty about the timing and intensity

of future changes. Rather than simply pulling up stakes, many households will hedge by shifting

more labor to urban markets (Dasgupta et al. 2016) and adjusting their property holdings if they

have any. Land transactions will evolve as climate change affects households differently and

transactors make different bets about future conditions.

Although empirical work on land market responses is well underway in developed countries, and

particularly the United States, data scarcity has hindered research in coastal developing areas. In

this paper, we utilize a new data source to address the issue at its “leading edge” in the

Sundarbans. Using a land transaction survey for 429 households and several climate-related

spatial databases, we analyze the adjustments of land prices and proprietors to changing

environmental conditions in the Indian Sundarbans. Our evidence is drawn from widely-

dispersed locales in three Blocks: Sagar Island, Kultali in the middle Sundarbans and Gosaba in

the eastern Sundarbans. Figure 1 provides a view of the area and locations of households

surveyed.

Figure 1: Sundarbans region - households sampled

3

The remainder of the

paper is organized as

follows. Section 2 provides an introduction to the Sundarbans region that highlights

demographic, economic and environmental forces that have shaped the current landscape. In

Section 3, we introduce our household survey dataset and our climate-related databases. Section

4 specifies, estimates and discusses the results of our econometric model, while Section 5

explores the implications for land pricing and household economic status in the Sundarbans.

Section 6 summarizes and concludes the paper.

2. Setting - The Indian Sundarbans

The Sundarbans is an intertidal estuarine zone comprising the deltaic plain formed by three major

rivers, the Ganga, Meghna and Brahmaputra. The Indian part of Sundarbans covers 9,630 km2,

bounded by the Ichhamati-Hariabhanga course in the east, the River Hugli in the west, the

Dampier–Hodges line in the north and the Bay of Bengal in the south. The region includes nearly

100 islands, of which half are inhabited and the remaining are demarcated as reserve forest. Its

intricate network of littoral ecosystems is studded with numerous water bodies, water channels,

rivers, creeks and islands supporting pockets of habitation.

About 4.5 million people currently reside in villages scattered across the Indian Sundarbans, and

their livelihoods are largely confined to agriculture and fishing (Danda, 2009). The rural residents

of the Sundarbans are among the poorest and most vulnerable people in India. Agricultural land

occupies 291,682 hectares (Danda, et al, 2015); production is largely rain-fed; and productivity is

substantially affected by salinity and weather hazards (Danda et al., 2011). While major storms

like cyclone Aila in 2009 have severe and obvious impacts, large coastal areas are also impacted

by tidal surges that may occur several times annually.

To investigate the implications of these conditions for land transactions, we have chosen to survey

households in three Blocks whose case histories reflect different change dynamics in the region.

Sagar and Gosaba CD Blocks are totally composed of islands and inaccessible by land transport.

4

Sagar is an important destination for religious and beach tourism. Gosaba has experienced some

growth in eco-tourism, which provides livelihood opportunities other than direct dependence on

the land or fishing (Hazra et al. 2010). In Kultali, increased settlement has been accompanied by

reduction of agricultural lands and orchards, while dense forests have increased marginally.

Gosaba, by contrast, has experienced reduction of dense forests and aquaculture as settlements,

agricultural lands and orchards have expanded. On Sagar Island, extensive land erosion has

occurred along with increases in settlement, aquaculture and dense forest, while agricultural lands

have receded.

3. Data

3.1 Household Land Transaction Survey

For this study, we have surveyed households in Sagar, Kultali and Gosaba (Figure 1). A list of

survey variables is included in the Appendix. The interview schedule was tested in a pilot study,

further modified and then administered for the full sample of households. Household selection

was based on plots of land identified for the study of transactions. Households which had made

land transactions were selected with the help of the village Amin and village elders.

Tables 1-4 tabulate sample transactions by price/hectare, parcel area, block and transaction type,

and transaction year. Unit land prices are calculated in INR 2017/hectare, using the World

Bank’s annual GDP deflator. The sample yields roughly balanced representation for eight price

categories, from prices below INR 1000/hectare (48 transactions) to prices above INR

100,000/hectare (61 transactions). Similarly, transaction parcels are distributed in approximate

balance across seven ranges, from parcels below 2 hectares (58 transactions) to parcels above 60

hectares (36 transactions). All three blocks have ample representation (Sagar 160 transactions;

Kultali 119; Gosaba 177), as do transaction types (purchases 268; sales 188). Transactions are

also widely spaced in time, with 88 in 2006, 30 in 2011, 62 in 2015 and 24 in 2016.

Table 1: Transaction prices (INR 2017/hectare)

Table 2: Transaction parcel areas (hectares)

Price Range

Upper Bound Count % Cum. %

1,000 48 10.53 10.53

2,500 54 11.84 22.37

5,000 72 15.79 38.16

10,000 79 17.32 55.48

25,000 75 16.45 71.93

50,000 37 8.11 80.04

100,000 30 6.58 86.62

100,000+ 61 13.38 100.0

Total 456 100.0

Area Range

Upper Bound Count % Cum. %

2 58 12.7 12.7

5 29 6.4 19.1

10 79 17.3 36.4

20 116 25.4 61.8

6

Table 3: Transaction Blocks and Types

Table 4: Transactions by year

3.1 Climate-Related Variables

Cyclone Strikes

Recurrent cyclonic storms in the Bay of Bengal inflict massive damage on the coastal regions of

Bangladesh and India. Extensive research has investigated the incidence, power and impacts of

cyclones in Bangladesh (Ali 1999; Dasgupta et al. 2014; Hoque 1992; Khalil 1992; Khalil 1993),

India (Mishra 2014; Srivastava et al. 2000) and the Bay of Bengal more generally (Bhaskar Rao

2001; Dash et al. 2004; Mandke and Bhide 2003; Mooley and Mohile 1983; Mooley 1980; Muni

Krishna 2009; Rao 2002; Yu and Wang 2009). Previous studies of coastal community

adaptation in the region have focused on responses to specific cyclones or generally-defined

coastal hazards (Khalil 1993; Khan et al. 2015; Mallick et al. 2011; Mallick and Vogt 2013;

Shameem and Momtaz 2014; Sultana and Mallick 2015).

In this paper, we extend the analysis using a georeferenced panel database of past cyclonic

storms (Bandyopadhyay et al. 2018) and the methodology of Dasgupta et al. (2019), which

computes cyclonic storm intensities in a multi-stage exercise. First, we assemble complete

georeferenced records for cyclonic storms in the region. For the period since 1970, we use

Block Purchase Sale Total

Sagar 110 50 160

Kultali 69 50 119

Gosaba 89 88 177

Total 268 188 456

Year Count % Cum. %

2006 88 19.3 19.30

2007 19 4.2 23.46

2008 26 5.7 29.17

2009 33 7.2 36.40

2010 26 5.7 42.11

2011 30 6.6 48.68

2012 42 9.2 57.89

2013 46 10.1 67.98

2014 60 13.2 81.14

2015 62 13.6 94.74

2016 24 5.3 100.0

Total 456 100.0

7

georeferenced track information on major cyclones striking Bangladesh from the Bangladesh

Meteorological Department (BMD). Equivalent information for the Indian coastal region comes

from the India Meteorological Department (IMD). We also use the IBTrACS data for the Indian

Ocean, provided by meteorological institutions in the region, for a few post-1960 cyclones that

are not included in the data available to us from BMD and IMD.

We ensure cross-source compatibility by using WMO standards for the two commonly-available

measures of cyclonic storm strength: maximum wind speed (measured in knots (kt)) and radial

distance from a storm’s center to its zone of maximum wind speed. We exclude all storms rated

as tropical depressions because their maximum wind speeds are below 34 kt. For each storm, we

compute the primary impact zone along its track as the area within the radius from centroid to

zone of maximum wind speed. 1 Using a methodology of the US National Hurricane Center

(USNHC 2018), we compute wind speed at each point after landfall as a function of wind speed

at landfall and elapsed time after landfall.2 We compute wind damage potential using a standard

exponential formulation (NOAAHRC 2018).3 We divide the historical storm data into three 15-

year periods: 1970-1984, 1985-1999 and 2000-2014. Then we compute an overall storm

intensity index using weights computed by Dasgupta et al. (2019) from regression analysis of

historical impacts on population displacement in the region. Figure 2 maps the estimated storm

intensities for the Indian Sundarbans region, which we will use for our econometric estimation

exercise in this paper.

1 For a complete discussion of database construction and impact area computation, see Bandyopadhyay et al. (2018). 2 In the USNHC model, the ratio [wind speed/wind speed at landfall] decays exponentially with time after landfall.

The absolute value of the exponential parameter is a positive function of wind speed at landfall (i.e., the rate of

decay is greater for storms with higher initial wind speeds). 3 In our computation, wind damage potential is proportional to the square of wind speed.

8

Figure 2: Indian Sundarbans - Storm intensity indicator for 1970 - 2014

Salinity

For this exercise, we employ an extensive spatial panel database of salinity readings for 342

monitoring stations in the Indian Sundarbans provided by WWF International (Figure 3). This

is an unbalanced panel, with many time series observations for some monitoring locations and

sparse observations for others. Table 5 tabulates the available observations by month and year.

Table 5: Sundarbans salinity monitor observations by month and year

Month 2012 2013 2014 2015

Jan 64 64

Feb 40 58 158 64

Mar 132 72 48

Apr 122 72 48

May 2 58 144 60

Jun 10 52 64 76

Jul 8 52 56 68

Aug 52 56 44

Sep 2 46 40 62

Oct 8 52 64 56

Nov 36 56 64 56

Dec 150 96 64 56

9

To fill in the panel, we estimate the following fixed-effects regression:

ln 𝑆𝑖𝑡 = 𝛽0 + ∑ 𝛽𝑗𝐷𝑆𝑗 + ∑ 𝛾𝑘𝐷𝑀𝑘 + 𝛿𝑦 + 휀𝑖𝑡

12

𝑘=2

𝑁

𝑗=1

where S = Salinity (ppt) at monitoring location i, period t

DS = Monitor dummy variable [1 for monitoring location j, 0 otherwise]

DM = Month dummy variable

y = year [2012, ..., 2015]

We use regression prediction to fill in the missing observations for the 342 monitoring stations in

all twelve months from 2012 to 2015. For the present exercise, we choose the 342 observations

from May, a peak month, in 2015.

Figure 3: Indian Sundarbans salinity monitor estimates, May 2015 (ppt)

4. Econometric Estimation

From the perspective of economic theory, the price of a land parcel is the present value of its

expected future rent flow at the prevailing discount rate. Climate-related factors have

significance in this context. Salinization should reduce land rent through its impact on soil

fertility (Dasgupta et al. 2017). Land rent should be lower in areas prone to cyclonic storm

damage, as well as areas susceptible to inundation and salinization from flooding and tidal flux.

Since land values are discounted rent flows, expectations about future inundation and

10

salinization are critical factors. The actual impacts of salinization, past storm intensity and

inundation risks will depend on their role in transactors’ formation of expectations about the

severity of future conditions.

Empirical work to date has focused principally on coastal areas of the United States, where storm

damage and inundation risk have been given more attention than salinization (West et al. 2001;

Bin, et al. 2010; McNamara and Keeler 2013; Yohe et al. 1996; Dachary-Bernard et al. 2019;

McAlpine and Porter 2018; Lichter and Felsenstein 2012). To our knowledge, this paper

represents the first such assessment for the coastal region of a developing country.

Household income affects the demand for land, and therefore its rental value, and land rental in

turn affects household income. Land price and household income are therefore jointly

determined. We have insufficient information for identification of structural equations in this

context, so we employ reduced-form estimation to investigate the impacts of salinization,

cyclonic storm damage and inundation risk on land prices and household incomes.

We specify the following estimation model:

(1) 𝑙𝑛 𝑋𝑖 = 𝛼0 + 𝛼1𝑙𝑛 𝐶𝑖 + 𝛼2 𝑆𝑖 + 𝛼3 𝐷𝑖 + 휀𝑖

Expectations: α1 > 0; α2, α3 < 0

where, for household i:

ln Xi = log income or land transaction (purchase or sale) price

ln Ci = Cyclonic inundation risk [log distance from coastline (see Figure 2)]

Si = Salinity

Di = Past and expected future wind damage from cyclonic storms Ɛi = Stochastic error, which may be subject to spatial autocorrelation

Land transactions are recorded by year from 2006 to 2016; we convert prices per hectare to INR

2017 using the India GDP deflator in the World Bank’s World Development Indicators. Income

(in INR 2018) is the sum of wages, product sales, land rental and out-migrant remittances

reported by survey households. For both land prices and incomes, we use log values to minimize

possible outlier effects. For Ci, we compute the distance from each household location to the

nearest point on a coastal polyline (Figure 2) constructed by the authors. We use the log value to

incorporate a non-linearity: The marginal protection benefit of distance from the coastline

should fall with distance. Si for a household is the reading for May 2015 from the nearest water

salinity monitor (Section 3.1; Figure 3). We use this as our proxy variable, following the finding

of Dasgupta et al. (2017) that land salinity in this coastal region is strongly predicted by

proximate water salinity. Di for a household is the nearest point in the spatial raster of past storm

severity indicators described in Section 3.1 and displayed in Figure 2.

4.1 Land Transactions Results

Table 6 reports our estimation results for land prices in recorded transactions. We include results

for OLS, GLS with standard errors adjusted for clusters of household groups, and a spatial HAC

11

estimator that incorporates both spatial and temporal autocorrelation. Since salinity and cyclone

strike intensity are correlated in the sample, we also include results for their first principal

component (PC1). We find the expected signs in all cases: Land transaction price increases with

distance from the coast at a decreasing marginal rate (elasticity < 1.0); decreases with salinity;

and decreases with cyclone strike intensity (CSI). Despite the sample correlation of salinity and

CSI, their independent covariation with land price is sufficient to yield consistently high

significance for OLS, Cluster SE and Spatial HAC. Their joint significance is highlighted by the

robust results for PC1. The magnitude of the estimated coastline distance effect is similar across

estimates, but significance is higher for the regressions that separate salinity and CSI.

4.2 Household Income Results

Table 7 reports our results for household incomes, including a term that identifies the transactor

(1 if seller; 0 if purchaser). As above, we report OLS, Cluster SE and Spatial HAC results for all

variables and PC1 for salinity and cyclone strike intensity. Here we find a coastline distance

effect that has both consistently high significance and an elasticity greater than the effect for land

prices. In this case, the consistently-robust PC1 results indicate the joint significance of salinity

and CSI, but their sample correlation is too high for their separate effects to be estimated with

confidence. We also find a large, significant adjustment for transactor identity: Sellers are

poorer than buyers and, since this is a logarithmic adjustment, the estimated income difference is

substantial.

5. Implications of the Results

Our econometric results suggest that salinization, inundation risk and cyclonic storm damage all

play significant roles in determining land prices and household incomes in the Indian

Sundarbans. However, their empirical importance hinges on actual impact magnitudes. To

address this question, we use our regression results to predict land prices and household incomes

for the most- and least-advantageous environmental conditions recorded in our database. Tables

8 and 9 display the two contrasting cases for distance from the coast (60.7 vs 0.7 km); salinity

(17.8 vs. 29.3 ppt) and the cyclone strike index (100 vs 137.5).

12

Table 6: Regression results - land price vs. climate-related variables

Dependent Variable: Log Land Price

OLS Cluster SE Spatial HAC OLS Cluster SE Spatial HAC

log Distance from Coast 0.282 0.282 0.282 0.204 0.204 0.204

(2.00)* (2.41)* (1.63) (1.58) (2.09)* (1.62)

Salinity -0.147 -0.147 -0.147

(4.01)** (2.36)* (2.68)**

Cyclone Strike Intensity -0.002 -0.002 -0.002

(3.09)** (2.00)* (2.20)*

PC1 [Salinity, -0.737 -0.737 -0.737

Cyclone Strike Intensity] (5.30)** (5.15)** (6.47)**

Constant 17.299 17.299 17.299 9.671 9.671 9.671

(10.71)** (10.44)** (13.35)** (39.59)** (30.67)** (36.77)**

Observations 456 456 456 456 456 456

R-squared 0.09 0.09 0.09 0.09

Absolute value of t statistics in parentheses

* significant at 5%; ** significant at 1%

13

Table 7: Regression results - income vs. climate-related variables

Dependent Variable: Log Income

OLS Cluster SE Spatial HAC OLS Cluster SE Spatial HAC

log Distance from Coast 0.385 0.385 0.385 0.371 0.371 0.371

(3.37)** (3.88)* (3.96)** (3.63)** (4.06)** (4.16)**

Salinity -0.049 -0.049 -0.049

(1.49) (1.38) (1.65)

Cyclone Strike Intensity -0.001 -0.001 -0.001

(1.61) (1.16) (1.79)

PC1 [Salinity, -0.299 -0.299 -0.299

Cyclone Strike Intensity] (2.62)** (2.08)* (2.25)*

Seller -0.501 -0.501 -0.501 -0.514 -0.514 -0.514

(2.43)* (1.53) (2.23)* (2.58)* (1.51) (2.34)*

Constant 13.452 13.452 13.452 10.423 10.423 10.423

(10.05)** (7.66)** (8.98)** (42.85)** (34.99)** (51.62)**

Obs 200 200 200 200 200 200

R-squared 0.08 0.08 0.08 0.08

Absolute value of t statistics in parentheses

* significant at 5%; ** significant at 1%

14

The associated predictions for land prices and incomes indicate that the environmental variables

have very large effects as a group. Predicted land prices under least and most favorable

environmental conditions are INR 751 and 96,359, respectively in the 7th and 87th percentiles of

the sample distribution. In the case of household incomes, we differentiate by transactor. For

sellers, incomes under the least and most favorable environmental conditions are INR 1,825 (5th

percentile) and 51,625 (86th percentile). For buyers, incomes are INR 2,990 (8th percentile)

and 84,580 (95th percentile). Overall, seller-buyer incomes under the least advantageous

environmental conditions are in the percentile range [5-8], while under the most advantageous

conditions they are in the percentile range [86-95].

Table 7: Predicted land prices under most and least favorable environmental conditions

Environmental

Advantage

Distance from

Coast (km)

Salinity

(ppt)

Cyclone

Strike Index

Land Price

INR (2017)/

Hectare

Most 60.7 17.8 100.0 96,359.3

Least 0.7 29.3 137.5 751.0

Table 8: Predicted incomes under most and least favorable environmental conditions

Environmental

Advantage

Distance from

Coast (km)

Salinity

(ppt)

Cyclone

Strike

Index

Total

Income

(INR 2018)

Most Buyer 60.7 17.8 100.0 84,579.5

Seller 60.7 17.8 100.0 51,625.1

Least Buyer 0.7 29.3 137.5 2,989.8

Seller 0.7 29.3 137.5 1,824.9

6. Summary and Conclusions

In this paper, we have used new household survey and environmental data to investigate the

relationships linking land prices, household incomes and climate-change-related factors in the

Indian Sundarbans. Although empirical work on land prices and climate change factors has

begun in developed countries, particularly the United States, we believe that this is the first such

exercise for the coastal region of a developing country.

Our household data are drawn from a georeferenced survey of land transactors that includes

information on plot sizes and prices, as well as the incomes and other characteristics of transactor

households. We quantify environmental conditions using two data sources: a georeferenced

panel of salinity measures drawn from 342 monitoring stations in the Indian Sundarbans region,

and a spatial raster database that incorporates information on all cyclonic storms that have struck

the region since 1970.

15

Using these data, we treat land prices and household incomes as jointly-determined variables in

an econometric model that incorporates the effects of salinization, cyclonic storm intensity, and

inundation risk from proximity to the coastline. We estimate reduced-form equations using

OLS, GLS with standard errors adjusted for spatially-grouped survey households and a spatial

HAC estimator that adjusts for spatial and temporal autocorrelation. For both land prices and

incomes, we find that the estimated parameters for the environmental variables have the expected

signs and generally-high levels of significance.

We use the regression results to predict land prices and household incomes for cases where

salinization, cyclonic storm intensity and inundation risk are at their least and most favorable

settings in the sample. We find very large impacts, with least and most favorable cases typically

yielding land prices and incomes below the 10th percentile and near the 90th percentile,

respectively.

We draw two broad conclusions from our results. First, current climate-change-related

conditions account for spatial differentials greater than an order of magnitude in Sundarbans land

prices and incomes. By implication, high-salinity, cyclone-prone areas near the coastline are

zones of poverty in the Sundarbans, while low-salinity areas far from the coastline with minor

cyclone histories are zones of relative affluence. Our second conclusion is driven by current

trends, which suggest that salinization, inundation risk, and cyclone strike frequency and

intensity will increase steadily as climate change proceeds. As this happens, our results suggest

that the most-affected areas will be abandoned by all but the poorest residents, who will be the

last to flee as the sea comes steadily inland.

16

References

Ali, A. 1999. Climate change impacts and adaptation assessment in Bangladesh. Climate

Research 12: 109–116.

Bandyopadhyay, S., S. Dasgupta, Z. Khan and D. Wheeler. 2018. Cyclonic Storm Landfalls in

Bangladesh, West Bengal and Odisha, 1877-2016: A Spatiotemporal Analysis. World Bank

Policy Research Working Paper (forthcoming).

Bandyoapdhyay, S. 1997. Coastal Erosion and Its Management in Sagar Island, South 24

Parganas, West Bengal, Indian Journal of Earth Sciences 24 (3-4): 51-69.

Bandyopadhyay, S. and M. Bandyopadhyay. 1996, Retrogradation of the western

Ganga–Brahmaputra delta, India and Bangladesh, possible reasons. Natl. Geogr.

31: 105–128.

Bhaskar Rao, D., C. Naidu and B. Rao. 2001. Trends and fluctuations of the cyclonic systems

over north Indian Ocean. Mausam 52: 37-46.

Bin, O., B. Poulter, C. Dumas and J. Whitehead. 2010. Measuring the Impact of Sea-Level Rise

on Coastal Real Estate: A Hedonic Property Model Approach. Journal of Regional Science,

51(4): 751-767.

Dachary-Bernard, J., H. Rey-Valette and B. Rulleau. 2019. Preferences among coastal and inland

residents relating to managed retreat: Influence of risk perception in acceptability of relocation

strategies, Journal of Environmental Management. 232:772-780.

Danda, A. 2009. in Ghosh, N., A. Danda, J Bandyopadhyay and S. Hazra (2015) Ecosystem

Rehabilitation and Adaptation to Global Warming and Climate Change in the ISD: An

Ecological-Economic Evaluation, WD Paper.

Danda, A., G. Sriskanthan, A. Ghosh, J. Bandyopadhyay and S. Hazra. 2011. Indian Sundarbans

Delta: A Vision (New Delhi: World Wide Fund for Nature-India).

Dasgupta, A. 2014. Land-Use Dynamics and Environmental Change in the Sundarbans:

Experiences from the Last Four Decades,International Journal of Geology, Earth &

Environmental Sciences; 4 (3): 1-8.

Dasgupta, S., M. Huq, Z.H. Khan, M.M.Z. Ahmed, N. Mukherjee, M. F. Khan and K. Pandey.

2014. Cyclones in a changing climate: the case of Bangladesh. Climate and Development 6(2).

Dasgupta, S., S. Bandyopadhyay and D. Wheeler. 2019. The Cyclone’s Shadow:

Historical Storm Impacts and Population Displacement in Bangladesh, West Bengal and Odisha.

World Bank (forthcoming).

17

Dasgupta, S., M. Hossain, M. Huq and D. Wheeler. 2016. Facing the Hungry Tide: Climate

Change, Livelihood Threats, and Household Responses in Coastal Bangladesh. Climate Change

Economics. 7(3): 1-25.

Dasgupta, S., M. Hossain, M. Huq and D. Wheeler. 2017. Climate Change, Salinization and

High-Yield Rice Production in Coastal BangladeshAgricultural and Resource Economics

Review, 47(1): 66-89.

Dash, S., J. Kumar and M. Shekhar. 2004. On the decreasing frequency of monsoon depressions

over the Indian region. Curr Science 86: 1404-1411.

Hazra, S. et al. 2002. Sea Level and associated changes in the Sundarbans, Science and Culture,

68(9-12): 309-321.

Hazra, S. et al. 2010. Temporal Change Detection (2001- 2008) Study of Sundarban, School of

Oceanographic Studies, Jadavpur University. pp. 1-127.

Hoq, M. 2007. An analysis of fisheries exploitation and management practices in Sundarbans

mangrove ecosystem, Bangladesh,Ocean & Coastal Management, 50: 411–427.

Hoque, M.M. 1992. Strategies and measures to reduce cyclone damage. In Cyclone disaster

management and regional/rural development planning: UNCRD-CIRDAP seminar, phase III,

Chittagong, Bangladesh (pp. 25–45). Nagoya: UNCRD, January 27–29.

Khalil, G. 1992. Cyclones and storm surges in Bangladesh: Some mitigative measures. Natural

Hazards 6: 11–24.

Khalil, G. 1993. The Catastrophic Cyclone of April 1991: Its Impact on the Economy of

Bangladesh. Nat. Hazards 8: 263–281.

Khan, M.., L. Bryceson, K. Kolivras, F. Faruque, M. Rahman and U. Haque, 2015. Natural

disasters and land-use/land-cover change in the southwest coastal areas of Bangladesh. Reg.

Environ. Change. 15: 241–250.

Lichter, M., and D. Felsenstein. 2012. Assessing the costs of sea-level rise and extreme flooding

at the local level: A GIS-based approach. Ocean & Coastal Management. 59:47-62,

Mallick, B., K. Rahaman and J. Vogt. 2011. Coastal livelihood and physical infrastructure in

Bangladesh after Cyclone Aila. Mitig. Adapt. Strateg. Glob. Chang. 16: 629–648.

Mallick, B. and J. Vogt. 2013. Population displacement after cyclone and its consequences:

Empirical evidence from coastal Bangladesh. Nat. Hazards 73: 191–212.

Mandke, S. and U. Bhide. 2003. A study of decreasing storm frequency over Bay of Bengal. J

Ind Geophys Union 7: 53-58.

18

McAlpine, S. and J. Porter. 2018. Estimating Recent Local Impacts of Sea-Level Rise on Current

Real-Estate Losses: A Housing Market Case Study in Miami-Dade, Florida, Population Research

and Policy Review. 37(6): 871-895.

McNamara, D. and A. Keeler. 2013. A coupled physical and economic model of the response of

coastal real estate to climate risk. Nature Climate Change, 3(6), 559–562.

Mishra, A. 2014. Temperature Rise and Trend of Cyclones over the Eastern Coastal Region of

India. J Earth Sci Clim Change 5: 227.

Mooley, D. and C. Mohile. 1983. A study of cyclonic storms incident in the different sections on

the coast around Bay of Bengal. Mausam 34: 139-152.

Mooley, D. 1980. Severe cyclonic storms in the Bay of Bengal, 1877-1977. Mon Wea Rev 108:

1647-1655.

Muni, K. 2009. Intensifying tropical cyclones over the North Indian Ocean during summer

monsoon – Global warming. Global Planetary Change 65: 12-16.

Rao, Y. 2002. The Bay of Bengal and tropical cyclones. Curr Sci 82: 379-381.

Shameem, M. and S. Momtaz. 2014. Vulnerability of rural livelihoods to multiple stressors: A

case study from the southwest coastal region of Bangladesh. Ocean Coast. Manag 23–38.

Srivastava, A., K. Ray and U. De. 2000. Trends in frequency of cyclonic disturbances and their

intensification over Indian seas. Mausam 51: 113-118.

Sultana, Z. and B. Mallick. 2015. Adaptation Strategies after Cyclone in Southwest Coastal

Bangladesh—Pro Poor Policy Choices. Am. J. Rural Dev 3: 24–33.

West, J., M. Small and H. Dowlatabadi. 2001. Storms, Investor Decisions, and the Economic

Impacts of Sea Level Rise. Climatic Change. 48(2–3): 317–342.

Yohe, G., J. Neumann, P. Marshall and H. Ameden. 1996.The economic cost of greenhouse-

induced sea-level rise for developed property in the United States. Climatic Change. 34(4): 387–

410.

Yu, J. and Y. Wang. 2009. Response of tropical cyclone potential intensity over the north Indian

Ocean to global warming. Geophy Res Let 36.

19

Appendix Table: Household Survey Variables

Variable Code Description

Sample_Id Sample Identity Number

Village 1.1 Village name (add, if something new)

Hamlet 1.2 hamlet name

HH_No 1.4 House number

Lat_HH 1.5 Latitude of HH

Long_HH 1.6 Longitude of HH

Head_of_HH 1.7 name of head of household

Informant 1.8 name of informant

Du_stay 1.9 Duration of stay in present location(in months)

Last_resi 1.10 Last place of residence (add if something new)

Place_of_Birth 1.11 Place of birth (add if something new)

R_ch_resi 1.12 Reason for change in residence (if place f birth & last

place of residence are not same)

HH_size 2.1 Household size

HH_Inds 2.2 principal industry of work

HH_Occu 2.3 principal occupation (add if something new)

HH_Type 2.4 household type

Prin_Inc 2.5 principal source of income

Soc_Gr 2.6 social group

Dwelling_unit 2.7 dwelling unit

House_type 2.8 type of structure

Drinking_water 2.9 principal source of drinking water

MGNREGS_card 2.10 whether the household has MGNREG job card

MGNREGS_days 2.11 Number of days of work during last 365 days

No_Outmigrated 4a.1 Number of HH members out-migrated

Out_Absence 4a.2 Average Duration of absence in last 365 days

(outmigration)

Out_Remittance 4a.3 Remittance received from out-migrant

Out_Mig_desti 4a.4 Out-migration destination (add if something new)

Out_Mig_reason 4a.5 Reason for out-migration (add if something new)

No_Inmigrated 4b.1 Number of HH members in-migrated

In_duration 4b.2 Average Duration of stayin present location (in-

migration)

In_Birth 4b.3 Place of Birth of in-migrant (add if something new)

In_last_resi 4b.4 Last Place of Residence of In-Migrant (add if

something new)

In_Reason 4b.5 Reason for in-migration (add if something new)

Land_Y_N 5.1 Whether household owns any land?

Land_Type 5.2 type of land owned (add if something new)

Agr_Land_O 5.a.4 Agricultural land owned (Kattah)

20


Agr_Land_Leasedin 5.a.5 Agricultural land Leased in(Kattah)

Agr_Land_otherwise 5.a.6 Agricultural Land neither owned nor leased-

in(Kattah)

Agr_Land_leasedout 5.a.7 Agricultural land Leased out(Kattah)

T_Agr_Land_possessed 5.a.3 Total Agricultural land possessed(Kattah)

Homestead_Land_O 5.b.4 Homestead land owned(Kattah)

Homestead_Land_LI 5.b.5 Homestead land Leased in(Kattah)

Homestead_Land_Otherwise 5.b.6 Homestead Land neither owned nor leased-in(Kattah)

Homestead_Land_Leasedout 5.b.7 Homestead land Leased out(Kattah)

T_Homestead_Land_possessed 5.b 3 Total Homestead land possessed(Kattah)

Pond_O 5.c.4 Pond land owned(Kattah)

Pond_Leased_in 5.c.5 Pond land Leased in(Kattah)

Pond_Otherwise 5.c.6 Pondl Land neither owned nor leased-in(Kattah)

Pond_leasedout 5.c.7 Pond land Leased out(Kattah)

T_Pond_possessed 5.c. 3 Total Pond land possessed(Kattah)

O_land_Owned 5.d.4 Any Other land owned(Kattah)

O_land_Leased_in 5.d.5 Any Other land Leased in(Kattah)

O_land_Otherwise 5.d.6 Any Other l Land neither owned nor leased-

in(Kattah)

O_land_out 5.d.7 Any Other land Leased out(Kattah)

O_land_Possed 5.d. 3 Total Any Other land possessed(Kattah)

Total_Land_possessed 5.3 Total Land Possessed

(Agr+Homestead+Pond+Other)(Kattah)

Agr_Production 5.8a Agriculture: Approximate quantum of production

during last 365 days (in Bosta)

Agr_price 5.9a Agriculture: Approximate price received/ estimated

Agr_Inc 5.10a Agriculture: Estimated income (in Rs)

Pond_Production 5.8c Pond: Approximate quantum of production during last

365 days

Pond_Price 5.9c Pond: Approximate price received/ estimated

Pond_Inc 5.10c Pond: Estimated income (in Rs)

Other_Production 5.8d Any other: Approximate quantum of production

during last 365 days

Other_Price 5.9d Any other: Approximate price received/ estimated

Other_Inc 5.10d Any other: Estimated income (in Rs)

Total_Inc_land 5.10 Total Income from land resource in Rs (Income from

Agr+Pnd)

No_Pcel 6. Total No of Pacels owned

Pcel_1_A 6.1 Parcel 1: Area(in Katha)

Pcel_1_D 6.2 Parcel 1:Distance from transport route(in m)

Pcel_1_Lat 6.3a Parcel 1:Latitude

Pcel_1_Long 6.3b Parcel 1:Longitude

21


Pcel_1_NIA 6.4 Parcel 1: Net Irrigated Area (NIA)

Pcel_1_Use 6.5 Parcel 1:Principal land-use of the plot (add if something

new)

Pcel_1_Crops 6.6 Parcel 1:No. Of crops raised

Pcel_1_E_Rent 6.7 Parcel 1:Expected/ actual rent if leased out

Pcel_1_Willingness 6.8 Parcel 1:Extent of Willingness to sell

Pcel_1_Reasons 6.9 Parcel 1:Reasons for Willingness to sell (add if

something new)







new)





something new)







new)





something new)




Pcel_4_long 6.3b Parcel 4:Longitude



new)




22



something new)

Land_sale_5yrs 7.1 How much land did you sell during last 5 years?(in

Katha)

Plot_1_Area_sold Plot 1:Area Sold(in Katha)

Plot_1_Time_sale Plot 1:Time of land sale (In month)

Plot_1_Loc_sold_plot Plot 1:Location of the plot sold

Plot_1_Price_rec Plot 1:Price Received(per Bigha)

Plot_1_Loc_buyer Plot 1:Place of residence of buyer (add if something new)

Plot_1_R_1_sale Plot 1:Reason-1 (in order of importance) for selling the

land (add if something new)


land


land

Plot_2_Area_sold Plot 2:Area Sold

Plot_2_Time_sale Plot 2:Time of land sale(In month)


Plot_2_Price_rec Plot 2:Price Received





land


land

Plot_3_Area_sold Plot 3:Area Sold

Plot_3_Time_sale Plot 3:Time of land sale(In month)


Plot_3_Price_rec Plot 3:Price Received





land


land

land_Sale_broker 7.3. Details of the broker involved in land sale mechanism

(add if something new)

land_Sale_broker_share 7.4. Share of the broker claimed in land transaction (in Rs.)

land_Sale_broker_process 7.5. Detail process involved in land transaction (add if

something new)

land_Sale_correct_price 7.6. Whether received correct price of land

23


sale_D_Cyclone 8.1 Land sale driver 1: Incidence of cyclones and other

weather disturbances have recently disrupted land based

activities-

Sae_D_salinity 8.2 Land sale driver 2: Increasing saline water intrusions

have destroyed soil and pond water quality. Hence want to

dispose them.

Sale_D_erosion 8.3 Land sale driver 3: Increasing coastal erosion is

threatening. Hence wish to lay-off the remaining land at

whatever price available.

Sale_D_frag 8.4 Land sale driver 4: Increased land fragmentation due to

succession has made the land holding un-viable & un-

economical for cultivation. It is better to dispose the tiny

holding.

Sale_D_price 8.5 Land Sale driver 5: Land prices are reasonably high. It

is therefore logical to sell the land and invest the money

elsewhere.

Sale_D_broker 8.6 Land sale driver 6: Mounting pressure from land

brokers.

Sale_D_Pol 8.7 Land sale driver 7: Conflicts over access to land are too

detrimental to land based activities. Hence it is worthwhile

to sell it off and take the money.

Sale_use_purpose_a 9.a.1 Purpose of use of money obtained from land sales


Sale_use_money_a 9.a.2 Usable amount (In Rs.)

Sale_use_purpose_b 9.b.1 Purpose of use of money obtained from land sales


Sale_use_money_b 9.b.2 Usable amount (In Rs.)

Sale_use_purpose_c 9.c.1 Purpose of use of money obtained from land sales


Sale_use_money_c 9.c.2 Usable amount (In Rs.)

Area_pur_5yrs1 10.1 How much land have you purchased during last 5

years?

Plot_1_Area_pur Plot 1:Area Sold(Katha)

Plot_1_Time_pur Plot 1:Time of land purchase(In month)

Plot_1_Loc_plot_pur Plot 1:Location of the plot purchased

Plot_1_Price_paid Plot 1:Price paid

Plot_1_Loc_seller Plot 1:Place of residence of seller (add if something new)

Plot_1_R_1_pur Plot 1:Reason-1 (in order of importance) for purchasing the



land


land

24










land


land









land


land

purchase_D_Remittance 11.1 Land purchase driver 1: Received reasonable amount

as remittance to invest in land purchasing.

purchase_D_Price_Soar 11.2 Land purchase driver 2: Land prices soaring. Hence

Trying to take part in land speculation

purchase_D_safer_area 11.3Land purchase driver 3: Desire to shift to safer

locations.

land_Purchase_broker_process 12.1 Detail process involved in land transaction: Purchase


land_Purchase_broker 12.2 Details of the broker involved in land sale mechanism


Wheather_Aquised 13.1 Wheather the land was aquised by Govt

Wheather_Relocated 13.2 Wheather relocated the household?

Time_of_Relo 13.3 Time of relocation in Months

Pur_of_Relocation 13.4 Purpose of relocation mention specifically

Compensation_amount 13.5 Amount received as compensation

discounting disaster: land markets and climate change in ...€¦ · the sundarbans landscape is...

Documents